Yersinia Enterocolitica Induces Epithelial Barrier

Yersinia enterocolitica induces epithelial barrier dysfunction through regional tight junction changes in colonic HT-29/B6 cell monolayers Nina A Hering1,2, Jan F Richter3, Susanne M Krug3, Dorothee Gu¨nzel3, Anja Fromm1,3, Erwin Bohn4, Rita Rosenthal3, Roland Bu¨cker2, Michael Fromm3, Hanno Troeger1 and Jo¨rg D Schulzke1,2

Yersinia enterocolitica is a common cause of acute gastroenteritis. This study aimed to clarify the mechanisms leading to barrier dysfunction and diarrhea. Exposure of human colonic HT-29/B6 cells to Y. enterocolitica resulted in a decrease in transepithelial resistance from 404±23 to 163±21 O cm2 (Po0.001) in parallel with an increase in mannitol (182 Da) and fluorescein (332 Da) permeability, whereas short circuit current did not change. This effect was time dependent, required the presence of living bacteria, could not be triggered by bacterial supernatants and was not due to Yersinia outer proteins. Concomitantly, Y. enterocolitica induced necrosis as indicated by an increase in lactate dehydrogenase-release, whereas epithelial apoptosis was not upregulated. Local changes in conductivity were detected by conductance scanning, indicating ‘leaky regions’ within the epithelium that were visualized by biotinylation and confocal microscopy. In these regions, claudin-3 and -4 and, especially claudin-8, were redistributed off the tight junction (TJ) into the cytoplasm. In addition, the expression of claudin-2, -3, -8, -10 and ZO-1 was diminished as quantified by immunoblotting. Moreover, we found claudin-8 to be regulated by the c-Jun N-terminal kinase, the inhibition of which attenuated the Y. enterocolitica-induced decrease in transepithelial resistance and restored claudin-8 protein level. In conclusion, barrier dysfunction in Y. enterocolitica infection is due to circumscribed epithelial TJ protein changes and necrotic cell loss, as a consequence of which leak flux diarrhea and antigen-uptake provoking extraintestinal arthritis may be triggered. Laboratory Investigation (2011) 91, 310–324; doi:10.1038/labinvest.2010.180; published online 18 October 2010

KEYWORDS: barrier dysfunction; c-Jun kinase; leaky region; necrosis; tight junction; Yersinia enterocolitica

Yersinia enterocolitica is a gram-negative, enteropathogenic invasin, which binds with high affinity to b1-integrins bacterium causing acute gastroenteritis, pseudoappendicitis exposed on M-cells located between the enterocytes. and mesenteric lymphadenitis. Depending on age and Invasin has an important role for internalization of yersiniae. immune status of the host, systemic manifestations, Y. enterocolitica is internalized by ‘zipper-like mechanisms’, including reactive arthritis, erythema nodosum, uveitis and passes the M-cell and reaches the Peyer’s patches, in which it septicaemia can be further complications of the infection.1,2 is able to proliferate and produce plasmid-encoded virulence Y. enterocolitica is a typical foodborne pathogen, which is factors. At 371C body temperature Y. enterocolitica is not primarily found in ground pork or refrigerated food, because motile and expresses a type-III-secretion system, as well as of its ability to multiply at temperatures close to 01C. Y. enterocolitica outer proteins (Yops), from its virulence Transmission occurs mainly through the fecal–oral route. plasmid (pYV). By translocating the Yop-effector proteins by Y. enterocolitica is known to express different sets of tem- the type-III-secretion system into the cytosol of the host perature-dependent virulence factors. A general infection cell Y. enterocolitica is assumed to suppress immune response model displays that ingested Y. enterocolitica, grown below (for review see Grassl et al3). The role of an enterotoxin that 301C are motile and express the outer membrane protein is exclusively produced below 301C in vitro is not clear so far,

1Department of Gastroenterology, Infectious Diseases and Rheumatology, Charite´ Campus Benjamin Franklin, Berlin, Germany; 2Department of General Medicine, Charite´, Campus Benjamin Franklin, Berlin, Germany; 3Institute of Clinical Physiology, Charite´, Campus Benjamin Franklin, Berlin, Germany and 4Institute of Medical Microbiology and Hygiene, University of Tu¨bingen, Tu¨bingen, Germany Correspondence: Professor Dr JD Schulzke, MD, Department of Gastroenterology, Infectious Diseases and Rheumatology, Charite´, Campus Benjamin Franklin, Hindenburgdamm 30, Berlin 12203, Germany. E-mail: [email protected] Received 4 March 2010; revised 23 August 2010; accepted 8 September 2010

310 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

but most data argue against a prominent role of a secretory Roswell Memorial Park Institute (RPMI) 1640 (PAA response by c-GMP from this toxin in vivo.4,5 However, Laboratories GmbH, Pasching, Austria), contained 10% fetal although Y. enterocolitica is a well-studied organism, only calf serum and 1% penicillin/streptomycin. Cells were in- little is known about the underlying mechanisms by cubated at 371C in an air atmosphere containing 5% CO2. which this pathogen causes barrier dysfunction leading to For electrophysiological measurements HT-29/B6 cells were diarrhea and possible antigen uptake. The impact of acute seeded on Millicell PCF filters (effective area 0.6 cm2;3mm Y. enterocolitica infection on intestinal barrier function was pores, Millicell PCF, Millipore, Schwalbach, Germany) and studied by our group in a mouse model in the past. Infection experiments were performed on confluent monolayers, giv- of mice with Y. enterocolitica O8 resulted in a decrease of ing transepithelial resistances (Rt) of at least 350 O cm2. epithelial resistance as measured by impedance spectroscopy, but showed unaltered Na þ -glucose co-transport or electro- Bacterial Growth Conditions And Infection genic ClÀ secretion.6 Bacterial strains used in this study are listed in Table 1. During the last two decades several enteric pathogens were Y. enterocolitica wild-type bacteria were grown in Luria- described to disturb the intestinal barrier in different ways, Bertani Broth (MP Biomedicals, Illkirch, France) at 271C including induction of local cell damage, eg by apoptosis, over night. Bacterial mutant strains were cultured in presence necrosis or focal leaks,7–9 stimulation of fluid and electrolyte of 10 mg/ml nalidixin acid or 25 mg/ml kanamycin. secretion, activation of inflammatory cascades, as well as For infection, a 1:8 dilution of the bacterial over-night disruption of the tight junction (TJ) of epithelial cells.10 culture was incubated for additional 2 h at 271C. Bacteria TJs are known as key determinants of barrier function.11,12 were harvested from log-phase cultures, washed and Localized between adjacent epithelial cells they limit the re-suspended in RPMI. For preparation of bacterial paracellular diffusion and regulate barrier function in response culture supernatant or lysate bacteria were grown in RPMI to different physiological requirements and stimuli.13,14 TJs are medium. Culture supernatant was filtrated using 0.2 mm composed of four different types of transmembranal proteins, ME DynaGard Syringe Tip Filter (Spectrum Laboratories, namely occluding,15 claudins,16 junctional adhesion mole- Rancho Dominguez, CA, USA) and concentrated using cules17 and tricellulin.18 These proteins interact with each Amicon Ultra-15 centrifugal filter units (Millipore), bacterial other, as well as with the actin cytoskeleton and membrane lysate was produced from 250 ml culture by means of associated cytosolic proteins (ZO-1, -2 and -3), performing the a hydraulic press (FRENCHPress, Thermo Spectronic, structural organisation of the junctional complex and enabling Cambridge, UK). regulatory inputs.19–21 In particular claudins, representing the Before infection confluent monolayers of HT-29/B6 cells major family of TJ proteins, are known to comprise sealing, were shifted to serum- and antibiotic-free RPMI medium. but in part also pore-forming properties.22 Disruption of TJ Cells were infected with 108 or 107 bacteria from the apical integrity can cause severe barrier dysfunctions as known from side equalling a multiplicity of infection (MOI) of 100 or 10. eg Vibrio cholerae zonula occludens toxin23 or Clostridium Incubation was carried out at 371C in an atmosphere of 24 perfringens enterotoxin. 95% O2 and 5% CO2. 2.5 h post infection (p.i.) 100 mg/ml In the present study, we aimed to characterize barrier gentamicin was added. dysfunction induced by Y. enterocolitica in a cell culture In parallel to infection with Y. enterocolitica strains, four model, in order to reveal diarrheal and barrier disturbing types of controls were performed: (i) without adding bac- mechanisms. teria, (ii) with addition of heat-inactivated bacteria (carried out at 951Cor561C for 15 min), (iii) lysates or supernatants MATERIALS AND METHODS and (iv) infection with E. coli K12. Cell Culture The cell line HT-29/B6, a subclone of the human colon Measurement of Transepithelial Resistance carcinoma cell line HT-29, is an appropriate model to Transepithelial resistance (Rt, O cm2) was determined with an study barrier function, reflected by (Rt).25 Culture medium ohmmeter (D Sorgenfrei, Institute of Clinical Physiology,

Table 1 Bacterial strains

Strain Description Reference or source

Ye O9 Y. enterocolitica serotype O9; clinical isolate provided by R Ignatius WA-P Y. enterocolitica serotype O8; clinical isolate; WA-314 pYVO8+ Heesemann and Laufs, 198326 WA-C Virulence plasmid cured derivative of WA-P, NalR Heesemann and Laufs, 198326 WA-P invÀ WA-P Dinv, NalR Ruckdeschel et al, 199627 WA-C invÀ Virulence plasmid cured derivative of WA-P Dinv, NalR Ruckdeschel et al, 199627

www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 311 Focal regulation of TJ by Y. enterocolitica NA Hering et al

Berlin, Germany) under sterile conditions at 371C. Data were The influence of calcium-chelation on Rt after infection corrected for the resistance of the empty filter. with Y. enterocolitica was measured after application of membrane permeable BAPTA-AM (Calbiochem, Darmstadt, Flux Measurement Germany) to the basolateral compartment. Confluent monolayers were shifted to RPMI medium and treated with 48 h after infection, HT-29/B6 cell monolayers were mounted À5 into Ussing-type chambers. Rt was determined by a com- 10 -M BAPTA-AM 1 h before infection. Infection procedure was carried out as described above. The 5 h p.i. BAPTA-AM puterized automatic clamp device (Fiebig Hard- & Software, t Berlin, Germany) and unidirectional flux-measurements was removed by basolateral medium exchange. R was mea- were performed from mucosal to serosal under short- sured 48 h. circuit conditions with [3H]-mannitol (Biotrend, Cologne, Germany), fluorescein and fluorescein isothiocyanate- Two-Path Impedance Spectroscopy Epithelial resistance (Repi), transcellular resistance (Rtrans) labelled dextran 4000 (Sigma-Aldrich, Schnelldorf, Germany) para according to previous protocols.28 and paracellular resistance (R ) of HT-29/B6 monolayers were measured by two-path impedance spectroscopy as Labeled dyes were added to the mucosal side and samples 31 were taken from the basolateral chamber at specific intervals. published recently. Briefly, filters were mounted in Ussing- Fluorescence was measured in a spectrofluorimeter (Spec- type impedance chambers and bathed with Ringer’s solution. traMaxGemini, Molecular Devices, Sunnyvale, CA, USA), Sinusoidal currents in frequencies from 1 to 65 kHz were [3H]-mannitol was analysed in a Tri-Carb 2100TR liquid provided by a programmable frequency response analyzer scintillation counter (Packard, Frankfurt, Germany) using (402, Beran Instruments, Devon, UK) in combination with Ultima Gold high flash-point liquid scintillation cocktail an electrochemical interface (1286, Solartron Schlumberger, (PerkinElmer LAS GmbH, Rodgau-Ju¨gesheim, Germany). Farnborough, UK). Additionally, fluxes of the paracellular marker, fluorescein were measured. Change of Rpara was Permeability was calculated from flux over concentration 2 þ difference. forced by Ca removal using ethylene glycol tetraacetic acid (EGTA) and the resulting impedance spectra and fluxes before and after chelating extracellular Ca2 þ were used for Electrogenic Chloride Secretion calculation. Chloride secretion was examined in Ussing-type chambers by À5 À5 the addition of 10 M of the blocker bumetanide or 10 M Conductance Scanning of the inductor forskolin (Sigma-Aldrich) to the basolateral To examine the spatial distribution of conductivity con- chamber and by quantifying the resulting changes in short ductance scanning was performed. Briefly, the cell monolayer circuit current (ISC). was placed horizontally in a chamber between two electrodes. Although an electric current passed through the monolayer, Measurement of Calcium Response the resulting electric field was detected on the apical side Intracellular calcium measurements were carried out as de- of the cell layer with a pair of microelectrodes. The tissue scribed previously29 using the Ca2 þ -sensitive dye fura-2AM conductivity could be calculated from the potential difference based on methods described by Grynkiewicz et al.30 Briefly, between these two electrodes. Local changes were detected by before each experiment HT-29/B6 cultured on coverslips an increase from basic conductivity when scanning parallel to were incubated with 2 Â 10À5 M fura-2AM (Sigma-Aldrich) the surface of the cell layer. For detailed description see in HEPES-Ringer’s solution (in mM: 151 Na þ ,5Kþ , 1.7 Troeger et al9 and Gitter et al.32 2 þ 2 þ À 2À À Ca , 0.9 Mg , 156.7 Cl , 0.9 SO4 ,1H2PO4 , 10 HEPES and 5 glucose) for 30 min at 371C. Then the coverslip was Lactate dehydrogenase (LDH) Release Assay placed into a perfusion chamber. After extracellular dyes As an indicator of cell disruption the LDH release from the cells were washed out HT-29/B6 monolayers were perfused with was measured according to Madara and Stafford.33 The LDH HEPES-Ringer’s solution containing vital or heat-inactivated content was measured in apical supernatants and in whole Y. enterocolitica (108 per ml) up to 1 h. Measurements were residual cells after detergent extraction with 2% Triton X-100 performed at 371C or room temperature. The excitation light for 20 min. The total LDH content of residual cells was set was generated by a xenon lamp (Osram XPO 75 W/2) filtered as 100% and LDH released into supernatants was calculated. by two rotating filters (6 per s) at 340 and 380 nm. The resulting fluorescence of fura-2AM was detected at 510 nm TUNEL-Staining for Apoptosis by a photomultiplier (Hamamatsu 928 SF) with consequent Cells were removed from filter supports with 0.5%. Trypsin- signal detection with an EPC-9 patch-clamp amplifier. EDTA (Invitrogen, Karlsruhe, Germany) at 371C. RPMI Changes in the 340/380 nm fluorescence ratio represented medium containing 10% fetal calf serum and 1% penicillin/ changes of intracellular calcium level. Calcium mobilization streptomycin was added and cells were separated with a from intracellular stores was stimulated by application of pipette tip. Cell suspensions were washed three times with 300 mM ATP (Sigma-Aldrich). phosphate-buffered saline (PBS) and cellular DNA was

312 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

stained with terminal deoxynucleotidyl transferase-mediated 1 mg/ml Sulfo-NHS-SS-Biotin (Cell Surface Protein Isolation deoxyuridine triphosphate nick-end labelling (TUNEL, Kit, Thermo Fisher Scientific, Bonn, Germany) was added In situ Cell Death Detection Kit, Fluorescein, Roche, Man- to the apical side. Binding was allowed for 30 min at 41C. nheim, Germany) according to the manufacturer’s protocol. Afterwards Sulfo-NHS-SS-biotin was removed and residual Samples were analysed by fluorescence activated cell sorting amine-reactivity was quenched by washing with ice cold (FACSCalibur, BD Biosciences, Heidelberg, Germany). The 100 mM glycin, followed by two further washing steps with apoptotic ratio was calculated as percentage of TUNEL- TRIS-buffered saline and PBS. Fixation was carried out with positive cells within total cells. As positive control apoptosis 2% paraformaldehyde at room temperature for 30 min. Cells was induced with a combination of 500 U/ml tumor necrosis were permeabilized with 0.05% Triton X-100 diluted in PBS factor a (TEBU, Offenbach, Germany) and 10 ng/ml and TJ proteins were immunostained with TJ-antibodies interleukin-13 (R&D systems, Wiesbaden, Germany) in (claudins 1:100, ZO-1 and occludin 1:200, Zymed) followed HT-29/B6.34 Both cytokines were added to the basolateral by incubation with Alexa Fluor 488 goat anti-mouse IgG or compartment and incubation was carried out at 371C for 48 h. Alexa Fluor 488 goat anti-rabbit IgG (1:500, Invitrogen) and Alexa Fluor 594 conjugated streptavidin (1:100, Invitrogen). Western Blot Biotinylation and changes in the TJ meshwork were visua- For protein expression analysis the western blot technique lized by confocal laser-scanning microscopy (Zeiss LSM was applied as described previously.35 Whole cell extracts 510META, Jena, Germany), using excitation wavelength of were generated with ice-cold lysis buffer, including 10 mM 543 nm for streptavidin-biotin and 488 nm for TJ proteins. Tris, pH 7.5, 150 mM NaCl, 0.5% Triton X-100, 0.1% SDS and complete protease inhibitor mixture (Roche). After Freeze Fracture Electron Microscopy incubation for 30–60 min on ice, insoluble material was Freeze fracture electron microscopy analysis was carried out removed by centrifugation at 15 000 g for 15 min at 41C. as described previously36 and is described in detail in the Protein concentrations were determined by Pierce BCA assay supplement. and 20–30 mg of whole cell extracts were loaded on poly- acrylamide gels. For immunodetection the following Quantitative Real-Time PCR (qRT-PCR) antibodies were used: anti-occludin and anti ZO-1 (1:2000, Total RNA was extracted from HT-29/B6 cells by Trizol Zymed, San Francisco, CA, USA), anti-tricellulin (custo- (Invitrogen) following the manufacturer’s protocol. For RT- mized by the group of O Huber, Institute of Clinical PCR High-Capacity cDNA Archive Kit (Applied Biosystems, Chemistry and Pathobiochemistry, Charite´, Germany) anti- Mannheim, Germany) was used. RT-PCR was carried claudin-1 to -5, -8 and -10 (1:1000, Zymed), MAPK Family out using TaqMan Gene Expression Assay no Hs00273282_s1 Antibody Sampler Kit and Phospho-MAPK Family Antibody for human claudin-8, Hs00265816_s1 for claudin-3 and Sampler Kit (Cell Signaling Technology, Danvers, MA, USA), Hs00268480_m1 for ZO-1. Human glyceraldehyde 3-phos- as well as anti-b-actin (1:5000, Sigma-Aldrich) as internal phate dehydrogenase (GAPDH, Applied Biosystems) served loading control. Densitometric quantification was carried out as endogenous control. Claudin-8- and GAPDH-cDNA were with AIDA quantification software (Raytest, Straubenhardt, quantified by VIC and FAM reporter dyes covalently attached Germany), values were normalized to b-actin. to the 50-terminal base of the probes. PCR was set up in duplicates, and threshold cycle values of the target genes were Biotinylation Assay normalized to the endogenous control. Differential expres- Visualization of ‘leaky regions’ in HT-29/B6 monolayers was sion was calculated according to the 2ÀDDCT method. performed 48 h p.i. using Sulfo-NHS-SS-Biotin, an amine- reactive biotinylation reagent, which binds to freely accessible Inhibition of Signaling Pathways primary amino groups of proteins. On tight monolayers only Confluent monolayers were shifted to RPMI medium con- proteins exposed at the apical cell surface are accessible for taining 2% fetal calf serum. Around 2 h before infection cells the apically added 606 Dalton Sulfo-NHS-biotin. In contrast, were treated with 10 mM c-Jun-terminal kinase inhibitor barrier defects enable the reagent to gain access also to the SP600125 (Sigma-Aldrich) or equal amounts of DMSO as basolateral space. Subsequent labelling of bound biotin using control from both sides. Alternatively, 2.5 mM JNKi-1 (Enzo a fluorescence-labelled streptavidin can visualize loci of life Sciences, Loerrach, Germany) was applied. Infection increased macromolecule passage. procedure and measurement of Rt were carried out as de- Biotinylation combined with immunostaining of TJ pro- scribed above. For all other inhibitors tested in this study teins and subsequent analyses by confocal scanning micro- please see Supplementary Table 1. scopy was carried out as follows. Medium was removed and monolayers resting on filter supports were washed and placed Statistics in PBS containing calcium and magnesium (PBS, PAA All data are expressed as means±s.e. of the mean (s.e.m.). Laboratories GmbH). In order to exclude internalization of Statistical analysis was performed using Student’s t-test, biotin, HT-29/B6 monolayers were cooled to 41C, before adjusted by Holm-Bonferroni correction for multiple

www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 313 Focal regulation of TJ by Y. enterocolitica NA Hering et al

comparisons. Po0.05 was considered significant (*Po0.05, Characterization of Barrier Impairment **Po0.01 and ***Po0.001). All subsequent analyses were performed on monolayers infected with 108 vital Y. enterocolitica O9 in parallel to untreated controls at 48 h p.i. To further analyze epithelial RESULTS barrier function monolayers were mounted into modified Infection With Y. enterocolitica Results in a Decrease of Ussing-type chambers. ISC was examined and unidirectional Transepithelial Resistance flux measurements with uncharged molecules of different Confluent monolayers of HT-29/B6 grown on permeable size were performed. 8 7 supports were infected with 10 or 10 Y. enterocolitica O9 Basal ISC was not affected by Y. enterocolitica (Figure 1b). from the apical side and were incubated at 371C. The bacteria Moreover, changes of ISC did not significantly differ from were killed by addition of gentamicin 2.5 h p.i. Barrier controls when chloride secretion was stimulated: Forskolin t integrity was monitored by measuring the R over 48 h at induced an increase in ISC of about 5.3±0.2-fold in 371C. In contrast to uninfected control monolayers, which controls and 6.4±0.6-fold in infected monolayers (n ¼ 3). exhibited only a slight decrease from initial resistance Inhibition of chloride secretion by bumetanide reduced ISC (85±6%; n ¼ 6), 108 Y. enterocolitica reduced Rt to 40±6% about 1.3±0.04-fold in controls and 1.2±0.05-fold in of initial resistance (Po0.01; n ¼ 6) at 48 h p.i. 107 bacteria Y. enterocolitica-treated cells. diminished Rt to 60±8% (Po0.05; n ¼ 6) suggesting a dose- In contrast, permeability for mannitol (182 Da, Stokes- dependence (Figure 1a). The E. coli K12 strain was tested as a radius 3.6 A˚ ) was increased 2.2-fold, namely from 2.3±0.2 in bacterial negative control and did not reveal any significant control to 5.3±0.3 Â 10À6 cm/s in Y. enterocolitica-treated differences compared with uninfected controls (71±2% in monolayers (Po0.0001; n ¼ 6). A 6-fold permeability K12-treated monolayers. vs 89±10% in controls; n ¼ 4). increase was measured for fluorescein (332 Da, 4.5 A˚ ).

120 6

100 5 ) 80 2 4

60 A/cm 3

* (

** SC 40 control I 2 MOI 10 20 1 (% of initial resistance)

t MOI 100 R 0 0 0 1020304050 control YeO9 time (h)

182 Da 332 Da 4000 Da mannitol fluorescein FITC-dextran 6 2.5 2.5 *** ** 2.0 2.0 cm/s)

-6 4 1.5 1.5

1.0 1.0 2 0.5 0.5 permeability (10 0 0.0 0.0 control YeO9 control YeO9 control YeO9

Figure 1 Characterization of barrier impairment. (a) HT-29/B6 monolayers were apically infected with 107 (multiplicity of infection (MOI) 10) or 108 Y. enterocolitica O9 (MOI 100) and incubated at 371C. 2.5 h p.i. bacteria were killed by the addition of 100 mg/ml gentamycin. Rt was monitored over 48 h and was dose and time dependent (n ¼ 6, *Po0.05, **Po0.01 vs control). (b) No changes in ISC were obtained (n ¼ 18). (c) Permeability of infected and control monolayers was determined by flux measurement in modified Ussing-type chambers. Y. enterocolitica increased permeability for 182 Da mannitol (n ¼ 6, ***Po0.001 vs control) and of 332 Da fluorescein (n ¼ 6, **Po0.01 vs control). For 4000 Da fluorescein isothiocyanate-Dextran layers from both conditions were almost impermeable.

314 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

The permeability was enhanced from 0.33±0.1 in control to distributed over the monolayer or represents a local phe- 1.79±0.3 Â 10À6 cm/s in infected monolayers (Po0.01; nomenon. In Figure 2a, the experimental ideas of con- n ¼ 6). However, when macromolecule permeation of 4 kDa ductance scanning and of biotin permeation across a ‘leaky fluorescein isothiocyanate-dextran (B13 A˚ ) was examined, region’ is depicted schematically. On each monolayer an area no significant changes were observed between controls of 1 mm2 was scanned in steps of 100 mm. As shown in (0.17±0.1 Â 10À6 cm/s; n ¼ 6) and Y. enterocolitica-infected Figure 2b, basal conductivity was unchanged in infected and cells (0.28±0.1 Â 10À6 cm/s; n ¼ 6) (Figure 1c). control monolayers (2.5±0.8 mS/cm2 after Y. enterocolitica infection vs 2.2±0.6 mS/cm2 in controls; n ¼ 7) and was Identification of Barrier Defect assumed to reflect intact regions of the monolayer. On top of The conductance scanning-technique was applied to elucidate this basal conductivity, local increases in conductance were 37 whether increased conductance occurred homogeneously observed (Gleak, mS). As described previously, such slight

a b c 10 4 intact area (A) 9 biotin conductance probe 8

preamp 3 7 )

2 6 S) μ ( intact (A) = leaky (B) = 2 5 basal leak leak G 4 3.29 conductivity conductivity apical G (mS/cm 3 1 2 "leaky regions" (B)

1 0.68 epithelial monolayer basal 0 0 control YeO9 control YeO9

d z z

z z

intact (A) intact (A)

leaky (B)

y y

30 μm 30 μm

x x control YeO9

Figure 2 Patchy distribution of barrier defects. (a) Cartoon illustrating both, conductance scanning and biotin permeation, in intact (A) and leaky (B) regions. For details see text. (b) Conductance scanning for regional resistance distribution was performed on infected and control monolayers. Basal conductivity of intact regions in infected monolayers was comparable to controls. Data are means±s.e.m. of n ¼ 7. (c) HT-29/B6 monolayers were scanned for sites exhibiting locally elevated conductances (controls and infected monolayers n ¼ 7 each). A dotted line was set to indicate a safe upper limit of local conductance peaks caused by apoptosis. Values above that line are regarded as ‘leaky regions’. Horizontal lines indicate means of the respective group. (d) A patchy distribution of ‘leaky regions’ was visualized by biotinylation assay (Z-axis stack) and subsequent confocal laser scanning microscopy. ‘Intact regions’ are marked as (A) and ‘leaky regions’ as (B).

www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 315 Focal regulation of TJ by Y. enterocolitica NA Hering et al

increases from basal conductivity were associated with reduced to 58±11% (Po0.01; n ¼ 6), 84±7% (Po0.05; apoptotic rosettes characterized by the enlargement of ad- n ¼ 6), 63±7% (Po0.01; n ¼ 6) and 70±4% (Po0.01; jacent cells visible in both controls and infected monolayers. n ¼ 6), respectively. ZO-1 was reduced to 52±4% As all leaks detected in controls were corresponding to (Po0.01; n ¼ 6), expression of occludin tricellulin, claudin-1, apoptotic events, an extrapolation line was drawn in safe -4 and -5 was unchanged (n ¼ 6) (Figures 3b and c). distance above these values (1.6 mS, Figure 2c). Localization of TJ proteins was characterized by confocal Apoptosis-assigned leaks exhibited a leak conductance of laser scanning microscopy. This was done in combination 0.68±0.07 mS (controls: 14 leaks and infected: 15 leaks). with the biotinylation assay for identification of ‘leaky However, in Y. enterocolitica-treated monolayers additional regions’ (Figure 4). Compared with confocal slicing regions of elevated leakiness were detected above the upper (approximately 0.5 mm) the height of the cells throughout the limit for apoptotic leaks (dotted line, Figure 2c) and assigned epithelial layer varies considerably. Therefore, the entire TJ as ‘leaky regions’. This ‘leaky region’-assigned leak con- meshwork staining was visualized in imaged fields by maxi- ductance was 3.29±0.56 mS (16 leaks) and thus five-fold mum projecting z-stacks on the xy-plane. In infected higher than that caused by apoptosis. monolayers, intact regions exhibiting no biotin passage Mean Rt of all control monolayers amounted to showed the typical TJ meshwork for ZO-1, occluding, as well 546±33 O Â cm2 equivalent to a conductivity of 1.9± as claudin-3, -4, -5 and -8 comparable to controls. In ‘leaky 2 0.1 mS/cm (Gcon). Infected layers had a mean total con- regions’ labelled by biotin-passage, no change in ZO-1 2 ductance of 3.5±0.3 mS/cm (GYe; Po0.001; n ¼ 7). Thus, (Figure 4a), occludin or claudin-5 was observed (data not 2 Gcon and GYe differed by 1.6 mS/cm (GD). From GD and shown). In contrast, the claudin-3, -4 and -8 patterns differed 2 Gleak a number of about 300 ‘leaky regions’/cm was esti- in ‘leaky regions’ from that of surrounding intact regions and mated (contributing to 46% of GYe). from controls. Claudin-4 (Figure 4b) and also claudin-3 (not Subsequently, it was possible to visualize highly permeable shown) were re-distributed as indicated by a loss of the regions by means of a biotinylation assay. A Sulfo-NHS- meshwork structure of the staining pattern and an enhanced biotin that binds to primary amino groups of accessible intracellular signal. However, changes in claudin-8 staining proteins was apically added to monolayers. After locating were even more pronounced. Claudin-8 appeared diffuse in bound biotin with fluorescence-labelled streptavidin it ‘leaky regions’ showing no typical TJ pattern any more became obvious that the intact HT-29/B6 monolayer did not (Figure 4c). In addition, mRNA levels for claudin-8, claudin- permit passage of biotin to the basolateral space at all, since 3 and ZO-1 were examined by qRT-PCR. As early as 6 h after only the apical surface was labelled in controls (Figure 2d). infection mRNA level of claudin-8 was reduced to 31±11% In contrast, Y. enterocolitica-infected monolayers exhibited (Po0.01; n ¼ 5), whereas the expression of ZO-1 mRNA was focal areas of intensive staining of cell boundaries (marked only slightly diminished to 79±4% (Po0.01; n ¼ 5). Level of with B in Figure 2d). These sites did not represent visible claudin-3 mRNA remained unchanged (n ¼ 5) (Figure 5). leaks but were identified as ‘leaky regions’, in which biotin could permeate from the apical side towards the basolateral TJ Strands Analyzed by Freeze Fracture Electron compartment. These ‘leaky regions’ appeared in different size Microscopy and randomly distributed over infected monolayers. Sur- Because TJs of ‘leaky regions’ were lost or at least critically rounding areas were not leaky and showed biotinylation only altered, to our experience such structures may not freeze- on the surface in accordance to controls (marked with A in fracture as complete strands. Thus the identified TJ strands in Figure 2d). infected cells may represent those of ‘intact regions’. As a result, no significant changes were observed between controls Paracellular Leakiness and Impaired TJs and Y. enterocolitica infection regarding number of horizon- Discrimination between paracellular and transcellular tally oriented strands, meshwork depth, density or occurrence resistance was achieved by two-path impedance spectroscopy. of strand breaks 420 nm. All strands were of the ‘‘con- The infection of HT-29/B6 resulted in an 8.5-fold reduction tinuous type’’, none of a ‘particle type’ (Supplementary of Rpara to 354±100 compared with 3012±496 O Â cm2 in Figure S1 and Supplementary Table S1). controls (Po0.01; n ¼ 6). Repi reflecting the total resistance of the monolayer was 3.8-fold reduced to 184±16 O Â cm2, Role of Prominent Virulence Factors whereas uninfected monolayers showed 704±21 O Â cm2 In order to elucidate the virulence factor of Y. enterocolitica (Po0.001; n ¼ 6). Transcellular resistance was not signifi- responsible for this barrier effect, the infection procedure was cantly changed (1147±477 in infected vs 978±181 O Â cm2 manipulated, seeking a crude classification of potential in control monolayers; n ¼ 6) (Figure 3a). virulence determinates. As the expression of different To further elucidate the paracellular leakiness, changes of Yersinia-virulence factors is temperature dependent, the TJ proteins were investigated. Claudin-1, -2, -3, -4, -5, -8 and bacteria were cultured at 27 or 371C before infection. How- -10, as well as ZO-1, occludin and tricellulin were quantified ever, the decrease of Rt in HT-29/B6 was independent of by immunoblotting. Levels of claudin-2, -3, -8 and -10 were bacterial culture temperature (data not shown). Moreover,

316 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

4000 175 Repi Rpara 150 3000 Rtrans 125 ) 2 100 * ·cm 2000 Ω 75 ** ** **

R ( ** 1000 50 (control set as 100%)

** densitometric intensity 25 *** 0 0 control YeO9 n

ZO-1

occludin claudin-1 claudin-2claudin-3claudin-4claudin-5 claudin-8 tricellulia + a1 claudin-10 control YeO9

claudin-1 -22 kDa

-22kDa claudin-2

claudin-3 -22 kDa

claudin-4 -22 kDa

claudin-5 -22 kDa

claudin-8 -22 kDa

claudin-10 -22 kDa

ZO-1 -220 kDa

occludin -56 kDa

tricellulin -63 kDa

-actin -42 kDa

Figure 3 Paracellular leakiness and impaired tight junctions. (a)Repi,Rtrans and Rpara were measured by two-path impedance spectroscopy. Repi, reflecting the total epithelial resistance, as well as the paracellular resistance Rpara were reduced in Y. enterocolitica-infected HT-29/B6 monolayers. Data are presented as means±s.e.m. of n ¼ 6(**Po0.01, ***Po0.001 vs control). (b) Western blot analyses revealed reduced protein levels of claudin-2 (n ¼ 9, **Po0.01), claudin-3 (n ¼ 9, *Po0.05), claudin-8 (n ¼ 12, **Po0.01) and claudin-10 (n ¼ 6, **Po0.01), as well as of ZO-1 (n ¼ 6, **Po0.01) in infected HT-29/B6 monolayers compared with control. (c) Densitometric quantification of TJ proteins after normalization with b-actin (n ¼ 6–12). only vital bacteria were able to cause the effect, as neither heat WA-P (serotype O8): (i) WA-C cured of the pYV, (ii) treatment of yersiniae (95 or 561C), nor Y. enterocolitica WA-P invÀ deficient for the chromosomal encoded invasion culture supernatants or bacterial lysates had any effect on Rt factor invasin and (iii) WA-C invÀ lacking both pYV and (Supplementary Figure S2). Also, YeO9-conditioned media, invasin. The wild-type strain WA-P reduced Rt to the produced by co-culturing of HT29/B6 and bacteria, had no same level as Y. enterocolitica O9. However, the infection of significant effect on Rt (data not shown) revealing the pre- HT-29/B6 with each of the three mutant-strains resulted sence of the epithelial cells was not necessary to trigger also in a significant drop of Rt (Figure 6). Although unin- production and/or secretion of a responsible Y. enterocolitica fected controls stabilized at 87 ± 3% (n ¼ 18), WA-C virulence factor. showed 51 ± 4% (Po0.001; n ¼ 18), WA-P invÀ 41±3% When the whole experiment was conducted over 48 h at (Po0.001; n ¼ 6) and WA-C invÀ 36±3% (Po0.001; 271C, no Rt decline occurred (88±13% after Y. enterocolitica n ¼ 6) of the initial resistance, 48 h p.i. Thus, both invasin- infection vs 93±15% in controls; n ¼ 6 each). deficient mutants caused even a slightly more pronounced In a further approach, mutant-strains were tested for their effect than the other Yersinia-strains (Po0.05, Po0.001). effect on Rt. HT-29/B6 monolayers were infected with Invasion-deficiency was proved in a gentamicin-protection three different mutant-strains derived from the wild-type assay (data not shown).

www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 317 Focal regulation of TJ by Y. enterocolitica NA Hering et al

Cytotoxicity and Apoptosis cells. In controls, a LDH-release of 1.1±0.1% was detected To examine, whether a loss of epithelial cells also contributes (n ¼ 12). In supernatants of infected cell monolayers, LDH to barrier dysfunction, the amount of necrotic and apoptotic concentrations were increased (3.8±0.8%, Po0.01; n ¼ 12). cells was determined. Y. enterocolitica-induced cytotoxicity However, the amount of LDH released did not correlate with was assessed by measuring the release of LDH from epithelial the decrease of Rt (Figure 7a). Mutant-strains WA-C, WA-P

a ZO-1 biotin merge

leaky intact leaky intact leaky intact

b claudin-4 biotin merge

leaky intact leaky intact leaky intact

Figure 4 Analyzes of tight junction proteins in leaky and intact regions by confocal laser scanning microscopy. The z-stack projections of infected monolayers displaying parallel biotinylation (red) and immunostaining for (a) ZO-1, (b) claudin-4 and (c) claudin-8 (green). ‘Leaky regions’ are characterized by an intense biotin permeation (red), which was absent in ‘intact regions’ that are characterized by lacking biotinylated cell boundaries.

318 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

c claudin-8 biotin merge

leaky intact leaky intact leaky intact

Figure 4 Continued.

1.2 100 90 1.0 80 Yersinia mutant strains ** 70 *** 0.8 60 *** 50 *** 0.6 *** 40 ** 30 0.4

(% of initial resistance) 20 t R

fold induction of mRNA 10 0.2 0 - - 0.0 WA-P WA-C control

ZO-1 WA-P inV WA-C inv t claudin-3 claudin-8 Figure 6 Effect on R after infection of HT-29/B6 monolayers with mutant strains. WA-C (PYV cured), WA-P invÀ (invasion-deficient) or WA-C invÀ Figure 5 Quantification of claudin-3 and -8 and ZO-1-mRNA-level. RNA was (PYV cured and invasion-deficient) derived from the Y. enterocolitica O8 prepared at 6 h p.i. from infected and control monolayers. mRNA expression wild-type (WAP) (n ¼ 6, ***P 0.001 vs control). of claudin-3, claudin-8 and ZO-1 were determined by qRT-PCR and GAPDH o was used for normalization. Data are means±s.e.m. of x-fold induction over controls (n ¼ 5, **Po0.01 compared with untreated control). invÀ and WA-C invÀ induced LDH-levels similar to wild- For comparison, tumor necrosis factora and interleukin-13 type yersiniae (data not shown). as a positive control induced an increase in apoptosis to Apoptosis induction was quantified in HT-29/B6 mono- about 17.8±0.7% (n ¼ 2). layers by TUNEL staining. Infection with Y. enterocolitica had no impact on epithelial apoptosis (Figure 7b). The number of Induction of Signaling Pathways apoptotic cells detected in infected monolayers was not Recently, it was published that the Y. enterocolitica heat-stable significantly different from that in controls (2.5±0.9% enterotoxin (Y-STa) stimulates intracellular inositol trispho- in infected monolayers vs 3.6±0.7% in controls; n ¼ 5). sphate (IP3) very quickly, which binds to the IP3 receptor www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 319 Focal regulation of TJ by Y. enterocolitica NA Hering et al

700 way, which in turn regulates TJ protein expression directly. control 600 A detailed description of all inhibitors tested is shown in YeO9 Supplementary Information, Supplementary Table S2. It is 500 interesting that only the c-Jun terminal kinase inhibitors ) 2 400 SP600125 or JNKi1 partially influenced the Y. enterocolitica- t ·cm induced decrease in R . All other inhibitors could not stop or Ω

( 300 t reduce the drop of resistance. R t 200 In Yersinia-infected monolayers treated with SP600125 R was reduced to 33±6.5% from initial resistance within 48 h 100 p.i. This was comparable to infected monolayers not treated 0 with SP600125 which amounted to 32±0.8% (n ¼ 11). t 0246810SP600125 diminished R to 61±0.7% vs control 78±3.3% LDH release (%) (Po0.01; n ¼ 11) masking a supposedly reduced decrease in Rt (Figure 8a). Thus, an other JNK-inhibitor, JNKi1, was 20.0 examined, which had no effect on Rt itself. JNKi1 reduced the t 17.5 Y. enterocolitica-induced decrease in R by 9±2%(Po0.01; n ¼ 4) (Figure 8b). 15.0 As shown in Figures 8c and d JNK-phosphorylation was 12.5 enhanced at 10 min after Y. enterocolitica infection (Po0.01; n ¼ 5), whereas other MAP kinases were not involved (Sup- 10.0 plementary Figure S3). In order to clarify, whether TJ protein 7.5 expression was affected by the inhibition of JNK protein level, ZO-1 and claudin-3, -4, -5 and -8 were examined by western apoptotic cells (%) 5.0 blot (Figure 8e). Inhibition of JNK resulted in a recovery of 2.5 claudin-8 expression, whereas all other TJ proteins tested were not affected. Treatment with SP600125 restored clau- 0 din-8 protein level to 90±6% compared with 66±10% in TNFa + control YeO9 untreated infected monolayers (P 0.05; n 5) and 96±9% IL-13 o ¼ in control monolayers treated with SP60025 (Figure 8f). Si- Figure 7 Y. enterocolitica-induced cytotoxicity. (a) Induction of necrosis in milar effects on claudin-8 expression were observed with HT-29/B6 monolayers was determined by measurement of LDH release. Furthermore, LDH release from controls and infected monolayers were JNKi1 (Supplementary Figure S4). However, biotinylation shown in correlation to the corresponding Rt values at 48 h p.i. (n ¼ 12). (b) and claudin-8 immunostaining of JNKi1-treated infected Induction of apoptosis was monitored by TUNEL-staining and subsequent monolayers revealed that JNK-inhibition could not influence fluorescence-activated cell sorting (n ¼ 5). HT-29/B6 monolayers treated the claudin-8 redistribution process within leaky regions with a combination of tumor necrosis factor a and interleukin-13 served as (data not shown). positive controls. Because the WA-C mutant strain decreased claudin-8 protein level as well at 48 h p.i. (Supplementary Figure S5), the observed effects on JNK were independent from the PYV. and mobilizes intracellular calcium in rat intestinal epithelial cells.38,39 Accordingly, we examined in how far a calcium- DISCUSSION response could be stimulated in HT-29/B6 by cultured Y. enterocolitica infection is associated with diarrhea and Y. enterocolitica using fura-2AM. As we expected a putative especially children can suffer from severe enterocolitis.41 calcium-response very quickly, we measured intracellular However, the inherent mechanisms and morphological calcium levels, about a duration, of 60 min after infection. As epithelial alterations are far from being clear. Until now, already described by Saha et al39 the potent IP3 agonist ATP40 only an epithelial barrier disturbance was identified as a was applied as positive control and induced a quick 4.2- prominent feature in a mouse model.6 This phenomenon fold±0.6 (n ¼ 3) rise in calcium from baseline levels could indeed contribute to diarrhea and may subsequently (221.6 nM±37, n ¼ 9) in HT-29/B6. However, when mono- lead to other pathological events frequently observed with layers were perfused with bacterial suspension (vital or heat- Y. enterocolitica infection, eg arthritis (cf. below). The inactivated yersiniae)at37or271C no calcium mobilization epithelial changes responsible for this barrier defect, however, occurred. Long-term effects were ruled out, as chelating have not been elucidated so far. intracellular calcium with BAPTA-AM before infection did not influence the course of Rt. Epithelial Changes in Y. enterocolitica Infection In the following, inhibitor studies were performed to ex- As known from previous studies HT-29/B6 is a suitable amine, whether Y. enterocolitica affected any signaling path- model for studying pathogen-induced barrier dysfunction.8,9

320 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

## 100 minutes after treatment with YeO9 100 control ** *** YeO9 control10 20 30 60 80 80 *** 54 kDa control P~JNK 60 60 YeO9 46 kDa 48 h p.i. 48 h p.i. t t 40 40 R R JNK 20 20 β-actin 42 kDa (% from initial resistance) (% from initial resistance) 0 0 SP600125 JNKi1

160 YeO9 YeO9 ** control control SP600125 YeO9 SP600125 140 ZO-1 120 claudin-3 100 claudin-4 P~JNK/JNK 80 claudin-5 (from 100% control)

60 claudin-8 06010 20 30 40 50 β-actin time (min)

SP600125 175 YeO9 150 YeO9+SP600125 125 # 100 ** * 75 ** * * 50

(control set as 100%) 25 densitometric intensity 0 1 ZO

claudin-3 claudin-4 claudin-5 claudin-8

Figure 8 Regulation of claudin-8 by JNK. HT-29/B6 monolayers were treated with JNK-inhibitors SP600125 or JNKi1 before infection. (a) In controls SP600125 caused a significant drop in Rt from initial resistance within 48 h, masking a supposedly reduced decrease in Rt in Y. enterocolitica-infected monolayers (n ¼ 11). (b) Application of JNKi1 had no effect on Rt in controls and an attenuation in Y. enterocolitica-induced Rt drop became obvious (Po0.01; n ¼ 4). (c) Western blot analyses showed enhanced JNK-phosphorylation at 10 min after infection. (d) Time course of JNK-phosphorylation in HT-29/B6 after Y. enterocolitica infection. Signals from all conditions were quantified by densitometry and the ratio of PBJNK and JNK was determined. Values from untreated controls were set as 100% and compared with signals from Y. enterocolitica stimulation. (e) Inhibition of JNK signalling restored claudin-8 protein level significantly, but had no impact on other TJ proteins probed (n ¼ 5). (f) Densitometric quantification of TJ proteins after normalization with b-actin (n ¼ 5or7;*Po0.05 and **Po0.01 compared with untreated control, #Po0.05 compared with inhibitor treatment).

In the present study we found that barrier function increased conductivity that could be detected by conductance was significantly changed in HT-29/B6 monolayers by scanning. Structurally, this regional loss in epithelial barrier Y. enterocolitica infection, indicated by resistance and flux function comprises changes in TJ protein assembly as measurements. As besides resistance only fluxes of para- visualized by confocal laser scanning microscopy. Within cellular markers with small hydrodynamic radii were these ‘leaky regions’ TJ proteins were re-distributed off the TJ enhanced (3.6 and 4.5 A˚ ), we propose Y. enterocolitica to into subapical intracellular compartments. trigger TJ regulation. This was also directly supported by data from two-path impedance spectroscopy. Role of Claudins in Y. enterocolitica Infection In addition to being distributed off the TJ, TJ proteins can Regional Leakiness in The HT-29/B6 Monolayers after also be downregulated in their expression level. In general, Y. enterocolitica Infection several of the 24 claudins have barrier function and elevate One very important aspect of the present study is the Rt. Barrier-forming claudins present in small or large intes- detection of focal, unequally distributed barrier defects sur- tine are especially claudin-1, -3, -4, -5 and -8.42 Within these, rounded by ‘intact regions’ within infected monolayers. claudin-8 is a well-established barrier former of the TJ.43 In Functionally, these ‘leaky regions’ are characterized by an addition to its re-distribution process, claudin-8 was found

www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 321 Focal regulation of TJ by Y. enterocolitica NA Hering et al

to be strongly reduced on protein- and mRNA-level in favour to be invasin induced.56 In contrast, Y. enterocolitica did not of expression regulation from the gene. Other claudins elevate the apoptotic rate in HT-29/B6 monolayers, but in- create cation selective pores with a concomitant decrease in creased the level of necrotic cells. As LDH-levels were not resistance, eg claudin-244 and -10b.45 However, claudin-2 is diminished in mutant strain WA-C, WA-P invÀ and WA-C not present in the intestine under physiological conditions invÀ exposed monolayers when compared with wild-type and is even slightly up-regulated during inflammation or yersiniae, cytotoxicity has to be assumed to be independent epithelial restitution. The role of claudin-10 for intestinal from invasin or Yop-proteins. Although necrotic cell death is barrier function is less well defined, as it was so far studied very likely to contribute to leakiness, we did not find a cor- only in the kidney and its functional role has to be assumed relation of LDH-release and Rt-decrease 48 h after infection. to depend on the specific environment of the other TJ pro- One possible explanation is that Y. enterocolitica-induced teins in the intestine. Thus, taken together, claudin-3, -4 and necrotic leaks are properly closed by surrounding cells. -8 which are reduced or/and re-distributed off the TJ by Otherwise, it may be possible that barrier dysfunction from Y. enterocolitica infection are the most prominent candidates Y. enterocolitica-infection is triggered by the induction of to contribute to the barrier dysfunction observed in the necrosis that could end up in proteolytic cleavage of TJ present study. This is in agreement with other investigations proteins in epithelial cells, a phenomenon that is known eg describing different enteric pathogens to cause barrier from conditions with induction of apoptosis.57 dysfunction by affecting claudins.23,24,46 Mechanisms of Interaction Role of ZO-1 It is interesting that the barrier function was only affected ZO-1 protein expression was substantially diminished in by viable bacteria and not by supernatants of cultured infected monolayers. This membrane-associated scaffolding Y. enterocolitica. Other bacteria as eg Vibrio cholerae secrete protein47 couples the major transmembranal barrier proteins toxins that disturb barrier function, namely the zonula oc- as claudins,48 occludin,49,50 tricellulin18 and JAM-A17,51 to cludens toxin.23 Also, production and release of hemolysins myosin-actin cytoskeleton.52 As ZO-1 is not a strand com- as in the case of E. coli O4 represent an example for barrier ponent it has no direct effect on transepithelial resistance, but modification by soluble factors.9 However, Y. enterocolitica it is not unlikely that a reduction of the ZO-1 anchor can supernatants or lysates were not effective. Accordingly, we do result in a disassembly of barrier relevant claudins from the not expect toxins or lipopolysaccharides to be involved. The TJ strands.8 role of the Yersinia heat-stable enterotoxin, which is only produced below 301C in vitro,5 is controversially discussed. TJ Ultrastructure by Freeze Fracture Electron Microscopy Saha et al38,39 showed Y-STa to stimulate IP3-level quickly To all of our experimental experience, it is unlikely that such resulting in calcium-mobilization by IP3R in vitro. We did critically altered cell contacts like the ‘leaky regions’ would not find vital Y. enterocolitica to stimulate a calcium-depen- yield freeze fractures fulfilling the morphological criteria of dent response in HT-29/B6 cells. intact strands. Therefore, we presume that the identified Furthermore, the two mutant strains in our study still strands of infected cells represent ‘intact regions’. These TJ exhibited the barrier-disturbing effect. Thus, the barrier effect strands, however, were not significantly different to controls is independent of both invasin-mediated internalization and in any parameter of our morphometric analysis. This finding expression of the Yersinia pYV. These results differ from is in full accord with the conductance scanning data, which mechanisms described for Yersinia pseudotuberculosis, a close yielded in ‘intact regions’ no significant difference in ion relative of Y. enterocolitica. Here, the plasmid encoded conductance between controls and Y. enterocolitica-infected effector protein YopE of Y. pseudotuberculosis was shown to cells (Figure 2b). interfere with F-actin re-arrangement and subsequently to affect TJ structure and function.58 However, Y. enterocolitica Necrosis and Apoptosis Induction by Y. enterocolitica is more frequently associated with diarrhea than its close It is established that the loss of enterocytes can contribute to relative. Therefore, it is not unlikely that both bacteriae differ the leakiness of intestinal epithelia, as eg apoptotic events in in their virulence factors for barrier perturbation. the epithelium can increase the contribution of paracellular Several studies revealed Y. enterocolitica to induce pro- pathways to the overall epithelial permeability.7,37 However, inflammatory host cell response or to counteract host defense Y. enterocolitica did not induce apoptosis but epithelial mechanisms by interfering with eukaryotic signal transduc- necrosis, indicated by an increase in LDH release. tion pathways. For example, Yersinia invasin activates p38 Y. enterocolitica-derived cytotoxicity has already been and JNK in epithelial cells, which, together with MEK1, investigated in different tissues. The pYV-encoded YopP mediate the transcriptional activation of the interleukin-8 protein is essential for the induction of apoptosis in mac- gene.59 YopP induces cell death by the inhibition of MAPKs, rophages53 and simultaneously induces caspase-dependent predominantly p38 and JNK, in dendritic cells but not in apoptotic and caspase-independent necrotic cell death in macrophages.60 In this study, we found Y. enterocolitica to dendritic cells.54,55 In Hep2 cells, apoptosis has been shown induce JNK-activation resulting in a strong decrease in

322 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org Focal regulation of TJ by Y. enterocolitica NA Hering et al

claudin-8 expression that was independent from Yops. As regions’ due to the appearance of local TJ protein changes, also invasin-deficient mutants caused a pronounced decrease characterized by TJ protein reduction and re-distribution of in Rt compared with wild-type strains, we do not expect the TJ. On the other hand, necrotic cell loss takes place and invasin to be responsible for the JNK-activation in this set- may contribute to the barrier loss. ting. Although JNK inhibition restored claudin-8 protein level in infected monolayers to normal, it had no influence on Supplementary Information accompanies the paper on the Laboratory claudin-8 delocalization within leaky regions. This may Investigation website (http://www.laboratoryinvestigation.org) explain the slight effect of JNK-inhibition on Rt. Moreover, Y. enterocolitica exerts a multimodal effect on epithelial ACKNOWLEDGEMENTS This work was supported by DFG Research Unit FOR 721. We thank Ralf monolayers. Consequently, it is not surprising that restitution Ignatius for kindly providing the clinical Y. enterocolitica O9 isolate. of only one single component could only partially restore Furthermore, we thank D. Sorgenfrei, S. Scho¨n and I.M. Lee for their barrier function. Modulation of TJ protein expression by excellent technical assistance. DFG Grant FOR 721. different signaling cascades has been described for different cell lines and conditions.61–63 However, so far it is not clear, DISCLOSURE/CONFLICT OF INTEREST which signaling cascades alter the expression of ZO-1 and The authors declare no conflict of interest. other claudins than claudin-8 in Y. enterocolitica infection. Moreover, the type of interaction remains speculative. A 1. Smego RA, Frean J, Koornhof HJ. Yersiniosis I: microbiological and clinicoepidemiological aspects of plague and non-plague Yersinia receptor-mediated interaction between vital bacteria and infections. Eur J Clin Microbiol Infect Dis 1999;18:1–15. epithelial cells might be hypothesized. Claudin-3 and -4 have 2. Koornhof HJ, Smego Jr RA, Nicol M. Yersiniosis. II: The pathogenesis of been described to serve as receptors for the Clostridium Yersinia infections. Eur J Clin Microbiol Infect Dis 1999;18:87–112. 3. Grassl GA, Bohn E, Muller Y, et al. Interaction of Yersinia enterocolitica perfringens enterotoxin, thereby inducing barrier dysfunc- with epithelial cells: invasin beyond invasion. Int J Med Microbiol 24 tion. Whether Y. enterocolitica directly interacts with TJ 2003;293:41–54. proteins remains uncertain. It seems more probable that the 4. Schiemann DA. An enterotoxin-negative strain of Yersinia enterocolitica serotype O:3 is capable of producing diarrhea in mice. patchy pattern of the barrier defect could be caused by a not Infect Immun 1981;32:571–574. equally distributed access of the bacteria to the epithelial 5. Mikulskis AV, Delor I, Thi VH, et al. Regulation of the Yersinia layer. This might depend on receptor exposure by single enterocolitica enterotoxin Yst gene. Influence of growth phase, temperature, osmolarity, pH and bacterial host factors. Mol Microbiol specialized cells within the epithelial monolayer or a dis- 1994;14:905–915. continuous protective mucus layer on the apical surface of 6. Gogarten W, Kockerling A, Fromm M, et al. Effect of acute Yersinia the cells. enterocolitica infection on intestinal barrier function in the mouse. Scand J Gastroenterol 1994;29:814–819. 7. Epple HJ, Schneider T, Troeger H, et al. Impairment of the intestinal Barrier-Disturbing Mechanisms of Y. enterocolitica barrier is evident in untreated but absent in suppressively treated HIV- Infection infected patients. Gut 2009;58:220–227. 8. Bucker R, Troeger H, Kleer J, et al. Arcobacter butzleri induces barrier A leak flux mechanism of diarrhea as a result of the impaired dysfunction in intestinal HT-29/B6 cells. J Infect Dis 2009;200:756–764. barrier function had been proposed by our group for 9. Troeger H, Richter JF, Beutin L, et al. Escherichia coli alpha-haemolysin Y. enterocolitica-infection.6 In support of this, in the present induces focal leaks in colonic epithelium: a novel mechanism of bacterial translocation. Cell Microbiol 2007;9:2530–2540. study on HT-29/B6 cells we identified the underlying me- 10. Berkes J, Viswanathan VK, Savkovic SD, et al. Intestinal epithelial chanisms of this barrier defect and confirmed the important responses to enteric pathogens: effects on the tight junction barrier, role of this pathomechanism. Other diarrheal mechanisms as ion transport, and inflammation. Gut 2003;52:439–451. 11. Anderson JM. Molecular structure of tight junctions and their role in active ion secretion or malabsorption were neither detected epithelial transport. News Physiol Sci 2001;16:126–130. in mouse small intestine nor in HT-29/B6 cells. This may also 12. Tsukita S, Furuse M, Itoh M. Multifunctional strands in tight junctions. explain why Y. enterocolitica is not that often detected as an Nat Rev Mol Cell Biol 2001;2:285–293. 13. Anderson JM, Van Itallie CM. Tight junctions and the molecular basis inducer of diarrhea and more frequently recognized in the for regulation of paracellular permeability. Am J Physiol 1995;269 clinic as a source of extraintestinal manifestations. In this (4 Part 1):G467–G475. context it should be mentioned that an infection-dependent 14. Madara JL. Regulation of the movement of solutes across tight junctions. Annu Rev Physiol 1998;60:143–159. increase in luminal antigen uptake resulting from intestinal 15. Furuse M, Hirase T, Itoh M, et al. Occludin: a novel integral membrane 64 barrier dysfunction has been proposed to trigger arthritis protein localizing at tight junctions. J Cell Biol 1993;123(6 Part 2): and inflammatory bowel diseases.65 1777–1788. 16. Furuse M, Fujita K, Hiiragi T, et al. Claudin-1 and -2: novel integral membrane proteins localizing at tight junctions with no sequence Conclusion similarity to occludin. J Cell Biol 1998;141:1539–1550. In the present study, we introduced mechanisms of barrier 17. Martin-Padura I, Lostaglio S, Schneemann M, et al. Junctional adhesion molecule, a novel member of the immunoglobulin superfamily that dysfunction relevant during Y. enterocolitica infection. This distributes at intercellular junctions and modulates monocyte pathogen exerts two different pathomechanisms, which both transmigration. J Cell Biol 1998;142:117–127. can cause leak flux diarrhea and seem not to depend on the 18. Ikenouchi J, Furuse M, Furuse K, et al. Tricellulin constitutes a novel barrier at tricellular contacts of epithelial cells. J Cell Biol 2005;171:939–945. typical Y. enterocolitica virulence factors, such as enterotoxin, 19. Turner JR. ‘Putting the squeeze’ on the tight junction: understanding invasins or Yops. On one hand, Y. enterocolitica induces ‘leaky cytoskeletal regulation. Semin Cell Dev Biol 2000;11:301–308.

www.laboratoryinvestigation.org | Laboratory Investigation | Volume 91 February 2011 323 Focal regulation of TJ by Y. enterocolitica NA Hering et al

20. Gonzalez-Mariscal L, Betanzos A, Avila-Flores A. MAGUK proteins: structure 45. Gu¨nzel D, Amasheh S, Pfaffenbach S, et al. Claudin-16 affects transcellular and role in the tight junction. Semin Cell Dev Biol 2000;11:315–324. Cl- secretion in MDCK cells. J Physiol 2009;587(Part 15):3777–3793. 21. Mitic LL, Van Itallie CM, Anderson JM. Molecular physiology and 46. Kohler H, Sakaguchi T, Hurley BP, et al. Salmonella enterica serovar pathophysiology of tight junctions I. Tight junction structure and Typhimurium regulates intercellular junction proteins and facilitates function: lessons from mutant animals and proteins. Am J Physiol transepithelial neutrophil and bacterial passage. Am J Physiol Gastrointest Liver Physiol 2000;279:G250–G254. Gastrointest Liver Physiol 2007;293:G178–G187. 22. Angelow S, Ahlstrom R, Yu AS. Biology of claudins. Am J Physiol Renal 47. Funke L, Dakoji S, Bredt DS. Membrane-associated guanylate kinases Physiol 2008;295:F867–F876. regulate adhesion and plasticity at cell junctions. Annu Rev Biochem 23. Fasano A, Baudry B, Pumplin DW, et al. Vibrio cholerae produces a 2005;74:219–245. second enterotoxin, which affects intestinal tight junctions. Proc Natl 48. Itoh M, Furuse M, Morita K, et al. Direct binding of three tight junction- Acad Sci USA 1991;88:5242–5246. associated MAGUKs, ZO-1, ZO-2, and ZO-3, with the COOH termini of 24. Sonoda N, Furuse M, Sasaki H, et al. Clostridium perfringens claudins. J Cell Biol 1999;147:1351–1363. enterotoxin fragment removes specific claudins from tight junction 49. Furuse M, Itoh M, Hirase T, et al. Direct association of occludin with strands: Evidence for direct involvement of claudins in tight junction ZO-1 and its possible involvement in the localization of occludin at barrier. J Cell Biol 1999;147:195–204. tight junctions. J Cell Biol 1994;127(6 Part 1):1617–1626. 25. Kreusel KM, Fromm M, Schulzke JD, et al. Cl- secretion in epithelial 50. Fanning AS, Jameson BJ, Jesaitis LA, et al. The tight junction monolayers of mucus-forming human colon cells (HT-29/B6). Am J protein ZO-1 establishes a link between the transmembrane protein Physiol 1991;261(4 Part 1):C574–C582. occludin and the actin cytoskeleton. J Biol Chem 1998;273: 26. Heesemann J, Laufs R. Construction of mobilizable Yersinia entero- 29745–29753. colitica virulence plasmid. J Bacteriol 1983;155:761–767. 51. Rehder D, Iden S, Nasdala I, et al. Junctional adhesion molecule-a 27. Ruckdeschel K, Roggenkamp A, Schubert S, et al. Differential contri- participates in the formation of apico-basal polarity through different bution of Yersinia enterocolitica virulence factors to evasion of domains. Exp Cell Res 2006;312:3389–3403. microbicidal action of neutrophils. Infect Immun 1996;64:724–733. 52. Hartsock A, Nelson WJ. Adherens and tight junctions: structure, 28. Epple HJ, Kreusel KM, Hanski C, et al. Differential stimulation of function and connections to the actin cytoskeleton. Biochim Biophys intestinal mucin secretion by cholera toxin and carbachol. Pflugers Acta 2008;1778:660–669. Arch 1997;433:638–647. 53. Mills SD, Boland A, Sory MP, et al. Yersinia enterocolitica induces 29. Rosenthal R, Choritz L, Zorn R, et al. Endothelin receptor B in trabecular apoptosis in macrophages by a process requiring functional type III meshwork. Exp Eye Res 2007;85:482–491. secretion and translocation mechanisms and involving YopP, 30. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ presumably acting as an effector protein. Proc Natl Acad Sci USA indicators with greatly improved fluorescence properties. J Biol Chem 1997;94:12638–12643. 1985;260:3440–3450. 54. Erfurth SE, Grobner S, Kramer U, et al. Yersinia enterocolitica 31. Krug SM, Fromm M, Gunzel D. Two-path impedance spectroscopy for induces apoptosis and inhibits surface molecule expression and measuring paracellular and transcellular epithelial resistance. Biophys J cytokine production in murine dendritic cells. Infect Immun 2004;72: 2009;97:2202–2211. 7045–7054. 32. Gitter AH, Bendfeldt K, Schulzke JD, et al. Leaks in the epithelial barrier 55. Grobner S, Autenrieth SE, Soldanova I, et al. Yersinia YopP-induced caused by spontaneous and TNF-alpha-induced single-cell apoptosis. apoptotic cell death in murine dendritic cells is partially independent Faseb J 2000;14:1749–1753. from action of caspases and exhibits necrosis-like features. Apoptosis 33. Madara JL, Stafford J. Interferon-gamma directly affects barrier function of 2006;11:1959–1968. cultured intestinal epithelial monolayers. J Clin Invest 1989;83:724–727. 56. Superti F, Pietrantoni A, Di Biase AM, et al. Inv-mediated apoptosis of 34. Heller F, Florian P, Bojarski C, et al. Interleukin-13 is the key effector Th2 epithelial cells infected with enteropathogenic Yersinia: a protective cytokine in ulcerative colitis that affects epithelial tight junctions, effect of lactoferrin. Res Microbiol 2005;156:728–737. apoptosis, and cell restitution. Gastroenterology 2005;129:550–564. 57. Bojarski C, Weiske J, Schoneberg T, et al. The specific fates of tight 35. Amasheh S, Schmidt T, Mahn M, et al. Contribution of claudin-5 to junction proteins in apoptotic epithelial cells. J Cell Sci 2004;117(Part barrier properties in tight junctions of epithelial cells. Cell Tissue Res 10):2097–2107. 2005;321:89–96. 58. Tafazoli F, Holmstrom A, Forsberg A, et al. Apically exposed, tight 36. Krug SM, Amasheh S, Richter JF, et al. Tricellulin forms a barrier to junction-associated beta1-integrins allow binding and YopE-mediated macromolecules in tricellular tight junctions without affecting ion perturbation of epithelial barriers by wild-type Yersinia bacteria. Infect permeability. Mol Biol Cell 2009;20:3713–3724. Immun 2000;68:5335–5343. 37. Bojarski C, Gitter AH, Bendfeldt K, et al. Permeability of human HT-29/ 59. Grassl GA, Kracht M, Wiedemann A, et al. Activation of NF-kappaB and B6 colonic epithelium as a function of apoptosis. J Physiol IL-8 by Yersinia enterocolitica invasin protein is conferred by 2001;535(Part 2):541–552. engagement of Rac1 and MAP kinase cascades. Cell Microbiol 38. Saha S, Gupta DD, Chakrabarti MK. Involvement of phospholipase C in 2003;5:957–971. Yersinia enterocolitica heat stable enterotoxin (Y-STa) mediated rise in 60. Adkins I, Schulz S, Borgmann S, et al. Differential roles of Yersinia outer intracellular calcium level in rat intestinal epithelial cells. Toxicon protein P-mediated inhibition of nuclear factor-kappa B in the 2005;45:361–367. induction of cell death in dendritic cells and macrophages. J Med 39. Saha S, Chowdhury P, Mazumdar A, et al. Role of Yersinia enterocolitica Microbiol 2008;57(Part 2):139–144. heat-stable enterotoxin (Y-STa) on differential regulation of nuclear 61. Mankertz J, Amasheh M, Krug SM, et al. TNFalpha up-regulates claudin-2 and cytosolic calcium signaling in rat intestinal epithelial cells. Cell Biol expression in epithelial HT-29/B6 cells via phosphatidylinositol-3-kinase Toxicol 2009;25:297–308. signaling. Cell Tissue Res 2009;336:67–77. 40. Berridge MJ. Inositol trisphosphate and calcium signalling. Nature 62. Carrozzino F, Pugnale P, Feraille E, et al. Inhibition of basal p38 or JNK 1993;361:315–325. activity enhances epithelial barrier function through differential 41. Marriott DJ, Taylor S, Dorman DC. Yersinia enterocolitica infection in modulation of claudin expression. Am J Physiol Cell Physiol 2009;297: children. Med J Aust 1985;143:489–492. C775–C787. 42. Markov AG, Veshnyakova A, Fromm M, et al. Segmental expression of 63. Naydenov NG, Hopkins AM, Ivanov AI. c-Jun N-terminal kinase claudin proteins correlates with tight junction barrier properties in rat mediates disassembly of apical junctions in model intestinal intestine. J Comp Physiol B 2010;180:591–598. epithelia. Cell Cycle 2009;8:2110–2121. 43. Yu AS, Enck AH, Lencer WI, et al. Claudin-8 expression in madin-darby 64. Sartor RB. Importance of intestinal mucosal immunity and luminal canine kidney cells augments the paracellular barrier to cation bacterial cell wall polymers in the aetiology of inflammatory joint permeation. J Biol Chem 2003;278:17350–17359. diseases. Baillieres Clin Rheumatol 1989;3:223–245. 44. Amasheh S, Meiri N, Gitter AH, et al. Claudin-2 expression induces 65. Saebo A, Vik E, Lange OJ, et al. Inflammatory bowel disease associated cation-selective channels in tight junctions of epithelial cells. J Cell Sci with Yersinia enterocolitica O:3 infection. Eur J Intern Med 2005;16: 2002;115(Part 24):4969–4976. 176–182.

324 Laboratory Investigation | Volume 91 February 2011 | www.laboratoryinvestigation.org